NEURONET: AI ASSISTANT FOR DETECTING AND MANAGING PARALYSIS IN PATIENTS

Early detection of facial and speech paralysis is critical for timely treatment and rehabilitation.Traditional diagnostic processes rely heavily on clinical examination and neuroimaging, which are often inaccessible in remote or resource-limited areas.This paper presents an AI-powered system capable of analyzing facial images and voice samples to detect early signs of paralysis using image preprocessing, audio encoding, and large language model (LLM)-based reasoning via API integration.The proposed system employs center-crop facial extraction, Base64 conversion, and structured medical prompting to achieve accurate detection.Experimental results demonstrate the system's effectiveness in identifying facial asymmetry, drooping patterns, and speech irregularities indicative of paralysis.The developed platform provides a lightweight, web-based solution suitable for telemedicine applications.